News about Rambus and the PS3

[marconelly!]

everything i've been hearing points to a 2006 ps3. If its late 2006 i think it will come with a blue ray . But it doesn't seem that the cell chips will be ready in time. I wonder if something is wrong at the plant.

Huh?

I think I've read a statement by Kutaragi himself, that Cell prototype will be ready by march 2004.

This news from bloomberg is just re-iterating some of the past month's news.

 

[Paul]

It all comes down to who are we going to trust.

Some article which states that the chip may not be ready until 2006 which is months old by now.

Or the recent news of Kutaragi, the main man in charge of Cell stating that Cell is on time and we are going to see a prototype soon.


Take your pick.

 

[jvd]

sorry . I meant for mass production. Enough volume to be put into a console . I'm sure the prototype will be done before the end of the year. I'm not sure about mass production . It seems like the plant is having problems

 

[Panajev2001a]

Weird, Toshiba has not posted any news on their site about the Oita fab not being ready for mass production by half of next year: they are still going by the previous schedule AFAIK...

 

[V3]

GSCube could have gone the micro-polygon approach: maybe not with the exact image quality obtained in off-line CG movies done with Renderman, but surely with decent rendering quality and decent frame-rate.

I don't doubt that, what I am asking, why don't they experiment with REYES pipeline on PS2 or GSCube.

I mean some people actually experiment with REYES implementation for real time purposes, on a lesser hardware too. I read about it last year, I think it was call Imagine or something along that line.

 

[Panajev2001a]

Well buy a PS2 Linux kit and let's start

 

[V3]

Well buy a PS2 Linux kit and let's start

Hehehe you can do it.

This guy probably can give you a head start, with his simple implementation of REYES. Giveout his source and several RIB file too.

http://dspace.dial.pipex.com/adrian.skilling/reyes/reyes.html

http://dspace.dial.pipex.com/adrian.skilling/reyes/reyes.tar

I've even seen some guy that try to do REYES using OpenGL, which was abit strange.

Get start'n

 

[Squeak]

I don't doubt that, what I am asking, why don't they experiment with REYES pipeline on PS2 or GSCube.

I mean some people actually experiment with REYES implementation for real time purposes, on a lesser hardware too. I read about it last year, I think it was call Imagine or something along that line.

Here it is: http://cva.stanford.edu/imagine/
Doesn’t exactly look less powerful than PS2, more like different but equal in power.

About the advantage of subpixel polys to do Stochastic AA you talked about earlier... It turns out that most REYES renderes today doesn’t use it, because most graphic chips now have has some sort of hardware AA build in, so they can save the resources for something else. The standard micropolygon size today in a REYES renderer is around one pixel.

 

[Panajev2001a]

Not in PRman, they specify 1/4th of a pixel IIRC

 

[Vince]

everything i've been hearing points to a 2006 ps3. If its late 2006 i think it will come with a blue ray . But it doesn't seem that the cell chips will be ready in time. I wonder if something is wrong at the plant .

I don't understand what people are thinking. The Cell Project was started in 2001 - you can't build a complex IC for 5+ years and not have feature-creep up the ass and not get leapfrogged by opponents. Or the Register and Cell in PS4 - I mean WTF, 12 years in R&D? This just doesn't happen...

 

[Paul]

you can't build a complex IC for 5+ years and not have feature-creep up the ass and not get leapfrogged by opponents.

Bingo.

 

[jvd]

everything i've been hearing points to a 2006 ps3. If its late 2006 i think it will come with a blue ray . But it doesn't seem that the cell chips will be ready in time. I wonder if something is wrong at the plant .

I sware, stupidity should be a crime.

So Jvd, what knowledge do you (that I and others lack) have that points toward this fact? Your great at talking, but it never adds up. Please... lay it all out, it's time to call this bluff.

Do realize that on the 2005 side I have slides from GDC '02 that state 2005. There is the Cell production in 2004 stated by Ken. There is the mass shipments of XDR beginning in mid-2005.

I just can't understand WTF people are thinking, or how they think. The Cell Project was started in 2001 - you can't build a complex IC for 5+ years and not have feature-creep up the ass and not get leapfrogged by opponents. Or the Register and Cell in PS4 - I mean WTF, 12 years in R&D? This just doesn't happen...

Hey i never said anything about xdr ram not being ready so you should stop making things up or confusing me with other people . Your dreams have to stop. Second of all anylists are saying the cell chip wont be in mass prduction till end of Fiscal year 2004 which is calander 2005. We have talked about this before. That is what is planed. From the anylist reports (which i tend to trust more than even ceo's of big companys. Nv30 taped out when ? ) I'm saying it seems like there are delays from the talks and in my opinion the ps3 might not launch till late 2006. Which has been said by others . Go read up past threads.

As for feature creep. Ever hear of the rampage ?

 

[Paul]

The XDR ram is going inside PS3, so find out when mass ammounts of the ram is going to ship and...

If we are going to trust Analyst's than no question Nvidia is going to have a GPU inside PS3, because there was that analyst that had that little article, well more like a few sentences.

There is no evidence to support that Cell/PS3 is not ontime, only speculation. Analyst's speculate, you know that.

If you want to call The Register, The Inquirer or Spong evidence than I just don't know what to say.. They are entertainment websites, NOT fact.

Yea there are TALKS, but look what the talks originate from... Designchain. EVERYONE of these articles you see out there about Cell/PS3 being late is just a rehash of this article.

Kutaragi just proved all this wrong, not to mention you have the ram shipping in huge quantaties in 2005.. You also have the Oita plant going into mass production 2004.

 

[Vince]

Hey i never said anything about xdr ram not being ready so you should stop making things up or confusing me with other people.

I'm not, what do you think they want XDR for? Maybe they'll use the mass production of it for... hmm

Your dreams have to stop. Second of all anylists are saying the cell chip wont be in mass prduction till end of Fiscal year 2004 which is calander 2005.

Link, please? And beyond that, since when do we trust analysts? They can't even keep the titles of games strait. Read the Register lately?

We have talked about this before. That is what is planed. From the anylist reports (which i tend to trust more than even ceo's of big companys. Nv30 taped out when ? ) I'm saying it seems like there are delays from the talks and in my opinion the ps3 might not launch till late 2006. Which has been said by others . Go read up past threads.

First of all, CEOs can't lie (just as nVidia's didn't) without their investors suing and brining in the SEC - this is common market sence. And I'm just saying you have no proof and have offered none here. Your like those people who follow the bandwagon just because - not noticing that all those articles all trace back to the Electronic Design Chain article that, if anything, is in support of Cell. It's people liek you or the analysts that don't keep up which just follow the precedent set by the Register or Sp0ng in saying it's delayed.

So, lets go - show me the empiracal evidence. Or, I can save you the trouble and tell you that you don't have shit.

As for feature creep. Ever hear of the rampage ?

You really have no idea who I am do you? Appearently you didnt visit here or TDFX's foolboard circa 2000... Please, don't attempt to lecture me on this.

 

[Panajev2001a]

till end of Fiscal year 2004 which is calander 2005.

BEEEP!

Wrong, try again...

Toshiba to Start Construction of Advanced Production Facility for 300mm Wafers for System LSI at Oita Operations

21 April, 2003

TOKYO--Toshiba Corporation today announced that construction of an advanced 300mm-wafer clean room for System LSI at its Oita Operations plant in Oita prefecture, Kyushu would start in June this year. The fab will start mass production in the latter half of first half of FY2004, and once it reaches full production will have a capacity of 12,500 300mm wafers a month. Approximately 40 billion yen will be invested in the new fab and its clean room in fiscal year 2003, as the first stage of a four-year, 200-billion yen scale project.

In December 2002, Toshiba announced its decision in principle to construct advanced 300mm-wafer fabs at its Oita Operations system LSI production base and its Yokkaichi Operations memory production base in Mie Prefecture. A four-year investment program from FY2003 will see approximately 350-billion yen channeled into the two new fabs. Today’s announcement on Oita is the first on specific plans under the program.

The new Oita fab will produce cutting-edge system LSIs, mainly microprocessors for broadband network applications. It will employ Toshiba’s embedded DRAM process technology and 65-nanometer process technology, and will transition to 45nm process in the future. By using advantages inherent in its world-leading embedded DRAM process technology, Toshiba is determined to remain a driving force in technological innovation and to provide customers with excellent products for broadband networks, including digital consumer products and mobile equipment.

Toshiba is now working with Sony Computer Entertainment Inc. (SCEI) on joint implementation of production facilities for the manufacture of SCEI’s products in Oita’s new clean room. The two companies will confirm details in due course, including the amount and timing of SCEI’s investment.

New Facility

Building Structure Two-floor, steel-framed building
Building ground area 24,100m2
Floor area 48,800m2
Clean room area 15,700m2
Start of construction June 2003
Completion January 2004
Start of mass production in the latter half of first half of FY2004 (planned)

Oita Operations

Location Oita City, Oita Prefecture
Established July 1970
General Manager Kuniaki Kumamaru
Employees approximately 3,000
Site area approximately 383,000m2
Building area approximately 198,000m

Mid FY2004 != End of FY2004...

 

[Vince]

Wow, nice post Paul.... well articulated, very nice. Me = teh impressed

 

[Paul]

*Bows*

 

[Squeak]

I just can't understand WTF people are thinking, or how they think. The Cell Project was started in 2001 - you can't build a complex IC for 5+ years and not have feature-creep up the ass and not get leapfrogged by opponents. Or the Register and Cell in PS4 - I mean WTF, 12 years in R&D? This just doesn't happen...

What feature-creep? We are not talking about a hardwired GPU here, where you need to be sure to include the newest buzzword technology.
On Cell, if you need a new feature you program it.
I should think the reverse is true, when you have such a relatively long time of development, it must be easier to think things through and make them elegant and "future proof".

 

[Panajev2001a]

Hey i never said anything about xdr ram not being ready so you should stop making things up or confusing me with other people.

I'm not, what do you think they want XDR for? Maybe they'll use the mass production of it for... hmm

Cut with the mis-information Vince... let me unmask you with the real application for XDR...

Rambus, Toshiba and Elpida Announce XDRâ„¢ DRAM, the World's Fastest Memory

3.2GHz DRAM offers 8x the bandwidth of today's best-of-class PC memory


Tokyo,Japan - July 10, 2003 - Rambus Inc. (Nasdaq: RMBS), a leading provider of chip-to-chip interface products and services, along with Toshiba and Elpida, today announced XDR? DRAM. XDR DRAM uses Rambus' XDR memory interface technology, formerly code-named Yellowstone. Running at 3.2GHz, XDR DRAM offers 8x the bandwidth of today's best-in-class PC memory. As Rambus announced earlier this year, Sony Corporation and Sony Computer Entertainment Inc. have licensed the XDR memory interface for utilization in future broadband applications with "Cell."

The XDR DRAM family has been architected to offer mainstream memory solutions for a broad range of applications. XDR DRAM is expected to initially serve the high-bandwidth needs of consumer, graphics, and networking applications, with eventual applicability for PC main memory, server and mobile systems when these applications require higher levels of bandwidth. XDR DRAM can provide the cost benefits of mainstream memory while still outperforming low-volume specialty DRAMs. XDR DRAM offers significant cost savings by providing the same system bandwidth as alternatives with fewer DRAM components, low-cost 4-layer PCBs, and inexpensive industry-standard packages.

Initially XDR DRAM will be offered at 3.2GHz with a roadmap to 6.4GHz and beyond, enabling memory system bandwidths up to 100GB/s. XDR DRAM will be available in multiple speed bins, device densities, and device widths. With densities ranging from 256Mb to 8Gb, and device widths ranging from x1 to x32, XDR DRAM satisfies the needs of both high-bandwidth and high-capacity systems. XDR memory's novel matrix topology allows point-to-point differential data interconnects to scale to multi-GHz speeds, while the bussed address and command signals allow a scalable range of memory system capacity supporting from one to 36 DRAM devices.

"Rambus is pleased that XDR DRAM is being supported by industry leaders. The XDR family provides a fresh approach to memory system design and will resolve bandwidth bottlenecks, enabling rich feature sets in next-generation broadband systems," said Laura Stark, vice president of the Memory Interface Division at Rambus. "With XDR memory, design engineers can successfully implement high performance systems at low system price points."

The XDR infrastructure, such as DRAM models, controller IO cells, clock generators, data sheets and system design guides, are available today for semiconductor and system design. From chip design to system integration and volume production, Rambus provides comprehensive services, support and a single point of contact for the entire memory system design to guarantee compatibility across multiple DRAM component vendors.

Toshiba and Elpida expect to begin shipping XDR DRAM in 2004, ramping to volume production in 2005. Additional information on XDR DRAM and the memory interface can be found at www.rambus.com/xdr.

About Rambus Inc.
Rambus is a leading provider of chip-to-chip interface products and services. The company's breakthrough technology and engineering expertise have helped leading chip and system companies to solve their challenging I/O problems and bring industry-leading products to market. Rambus' interface solutions can be found in numerous computing, consumer electronic and networking products. Additional information is available at www.rambus.com.

About Elpida Memory, Inc.
Elpida Memory, Inc. is a manufacturer of Dynamic Random Access Memory (DRAM) with headquarters based in Tokyo, Japan, and sales and marketing operations located in Japan, North America, Europe and Asia. Elpida offers a broad range of leading-edge DRAM products. Elpida is a joint venture company formed by NEC and Hitachi on December 20, 1999. Elpida has been in operation since April 2000. For more information, visit www.elpida.com.

About Toshiba Corporation
Toshiba Corporation is a leader in the development and manufacture of electronic devices and components, information and communication systems, consumer products and power systems. The company's ability to integrate wide-ranging capabilities, from hardware to software and innovative services, assure its position as an innovator in diverse fields and many businesses. In semiconductors, Toshiba continues to build on its world-class position in NAND flash memories, analog devices and discrete devices and to promote its leadership in the fast growing system-on-chip market. Toshiba has approximately 166,000 employees worldwide and annual sales of over US$47 billion.

Elpida Quote:
"Keeping up with ever-increasing bandwidth requirements is a critical element of memory system design and Rambus continues to demonstrate that they have the technical expertise to meet this challenge," said Hidemori Inukai, Chief Marketing Officer for Elpida Memory, Inc. "XDR DRAM has the potential to become a primary solution for many markets, especially high-end consumer and graphics systems."

Toshiba Quote:
"XDR DRAM ushers in a whole new era of advanced memory system design that addresses the needs of the bandwidth-intensive consumer marketplace," said Shozo Saito, Technology Executive of Toshiba Corporation Semiconductor Company. "We are confident that our long-standing partnership with Rambus will continue to equip our customers with leading DRAM products.

Rambus, RDRAM, and the Rambus logo are trademarks or registered trademarks of Rambus Inc. Rambus and other parties may have trademark rights in other terms used herein.

###

This release contains forward-looking statements that are made pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. Forward-looking statements include statements about the adoption rate of XDR DRAM into and suitability for the consumer, graphics, networking, PC main memory, server and mobile markets, as well as the dates of when XDR DRAMs are expected to ship into their respective markets and ramp into volume production. Additional forward-looking statements include the ability of XDR DRAM to provide cost benefits while still outperforming low-volume specialty DRAMs. These forward-looking statements are subject to risks and uncertainties, which could cause actual results to differ materially from those projected. Those risks include the possibility of inadequate shipments of Rambus XDR memory devices and controllers for the specified markets, the market response to these products, the continued deterioration in the DRAM market, any delay in the development of Rambus-based products by licensees, any delay in the development and shipment of new Rambus products, any delay in the development and shipment of products compatible with Rambus products, a strong response of the market to competing technology, a failure to sign new contracts or maintain existing contracts for XDR memory, adverse litigation decisions and other factors that are described in our SEC filings including our 10-K and 10-Qs.

Contacts:
Linda Ashmore
Rambus Public Relations
650-947-5411
lashmore@rambus.com

Natalie Kessler
The Hoffman Agency
408-975-3032
nkessler@hoffman.com

Tammy Lee
Lee Communications - PR for Elpida Memory (USA) Inc.
877-677-9533
tammy@leecommunications.com

Hiroshi Saito
Elpida Memory, Inc. (Japan)
+81-(0)3-3281-1604
saito-hiroshi@elpida.com

Peter Westerdorf
Elpida Memory, Inc. (Europe)
0049 (0)211 23945120
peter.westerdorf@elpida.com

Makoto Yasuda
Corporate Communications Office (Toshiba)
+ (81) 3 3457 2105
press@toshiba.co.jp

Rambus Signs Technology License Agreements With Sony, Sony Computer Entertainment and Toshiba

Two new interfaces selected for logic-to-memory and logic-to-logic connectivity

Rambus Inc. (Nasdaq:RMBS), the leader in ultra high-speed interface technology, today announced new agreements with Sony Corporation (Sony), Sony Computer Entertainment Inc. (SCEI) and Toshiba Corporation (Toshiba) for the license and utilization of two new high-speed interfaces, codenamed "Yellowstone" and "Redwood." Offering unparalleled competitive advantages, these two interfaces are expected to be utilized for future broadband applications with "Cell."

The impact on Rambus' financials will be discussed during a conference call and webcast on January 6, 2003 at 1:30 p.m. Pacific Standard Time. The specific terms of the agreements are confidential.

Currently at 3.2GHz data rates, with a roadmap to higher performance, "Yellowstone" is much faster than the best available DDR memories. "Yellowstone" offers high performance in memory signaling while optimizing system cost through pin-count reduction and support for high volume PCBs and packages.

"Redwood," the ultra high-speed parallel interface between multiple chips, delivers a data rate about ten times faster than the latest processor busses. It maintains lower latency and lower power consumption than current solutions, while keeping high productivity and cost efficiency.

"The use of Direct Rambus technology in PlayStation®2 was essential for its performance," said Ken Kutaragi, president and chief executive officer of Sony Computer Entertainment Inc. "Rambus is and will be the key player in the ultra high-speed interface technology. This enables us to create a wide range of applications and platforms from high-end systems to digital consumer electronics products within Sony Group."

"We recognize Rambus as the premier provider of high speed interface technology. We have already decided to integrate Rambus' interface technology into our next-generation high-value added DRAM, and we have now extended our partnership to the logic interface. These technologies will support us in delivering effective solutions to next-generation systems that require high-speed processing of large graphics and audio data," said Takeshi Nakagawa, corporate senior vice president of Toshiba Corporation and president and chief executive officer of Toshiba Corporation Semiconductor Company.

"We have had long and mutually beneficial relationships with Sony, Sony Computer Entertainment and Toshiba," said Geoff Tate, chief executive officer at Rambus. "Rambus' objective is to produce innovative solutions that will benefit our semiconductor and system partners. We are pleased that our ultra high-speed logic-to-memory and logic-to-logic solutions are key technologies to produce a wide range of future systems."
About Rambus Inc.


Rambus is a leading provider of chip-to-chip interface products and services. The company's breakthrough technology and engineering expertise have helped leading chip and system companies to solve their challenging interface problems and bring industry-leading products to market. Rambus interface solutions can be found in hundreds of computing, consumer electronic and networking products. Additional information is available at www.rambus.com.

This press release contains forward-looking statements. These statements are based on current expectations, estimates and projections about the Company's industry, management's beliefs, and certain assumptions made by the Company's management. You can identify these and other forward-looking statements by the use of words such as "may," "will," "should," "expects," "plans," "anticipates," "believes," "estimates," "predicts," "intends," "potential," "continue" or the negative of such terms, or other comparable terminology. Forward-looking statements also include the assumptions underlying or relating to the foregoing statements. Actual results could differ materially from those anticipated in these forward-looking statements as a result of various factors, including those identified in the Company's recent filings with the Securities and Exchange Commission, including its recently filed Form 10-Q, and also including the uncertainty of new technologies; and the uncertainty regarding the technical and market demands for such technologies. All forward-looking statements included in this press release are based on information available to Rambus on the date hereof. Rambus assumes no obligation to update any forward-looking statements.

Rambus is a registered trademark of Rambus Inc. PlayStation is a registered trademark of Sony Computer Entertainment Inc. Other trademarks that may be mentioned in this release are the intellectual property of their respective owners.

Contacts:
Linda Ashmore
Rambus Public Relations
650-947-5411
lashmore@rambus.com

Heather Carlson
The Hoffman Agency
408-975-3034
hcarlson@hoffman.com

 

[Vince]

On Cell, if you need a new feature you program it.
I should think the reverse is true, when you have such a relatively long time of development, it must be easier to think things through and make them elegant and "future proof".

You also need to design to a specific mask size with a specific transistor limit, thermal tolerances, et al. For example, an IBM fellow whose one of the 5 men from IBM who jointly designed the Cell architecture with the S/T teams previously stated in a paper that a temporal slip of 1 month costs 4% in relative preformance. 6 Months is roughly 26%, 18 months is a full 100% loss in market relative preformance.

And people like Jvd here are saying a 1-2 year slip due to problems at the "plant."

PS. Thanks Pana, clearification is good - can never be too sure...